π’ Open to Senior ML/AI Engineering roles & AI consulting opportunities
π Bengaluru, India β’ β‘ Usually responds within 24h β’ π ananttripathi.in
Profile Views Followers
Connect
LinkedIn X Portfolio Kaggle HuggingFace
Certifications
Stanford ML IBM GenAI Google GenAI
Project Links
Tourism Airline HR Bot Pneumonia Detection Medical RAG Resume Analyzer Roadmap DSCode
"Building intelligent systems that don't just predict the futureβthey optimize it."
I'm a Senior ML & AI Engineer with 5+ years of experience building production-grade AI solutions across LLMs, optimization, and predictive analytics. Currently leading data science initiatives at Axtria β Ingenious Insights while pursuing 3 advanced AI/ML programs simultaneously (UT Austin, IIIT Bangalore, Deakin University).
What I Do:
- π§ Build and deploy GenAI applications using LLMs, RAG systems, and Azure OpenAI
- π― Architect marketing mix optimization platforms serving Fortune 500 pharma clients (Bayer, Merck, Novartis, Janssen)
- π Design scalable MLOps pipelines with Docker, MLflow, FastAPI, and CI/CD automation
- π Lead cross-functional teams delivering 25+ data science projects with measurable business impact
- π Mentor engineers and train 70+ professionals in ML, Python, SQL, and optimization strategies
- ποΈ Own 10+ product capabilities from design to deployment with enterprise-scale impact
Career Highlights:
- π 4 promotions in 5 years: Analyst β Associate β Senior Associate β Project Leader β Manager
- β‘ 98-100% error-free delivery rate across production releases
- π― 95%+ on-time delivery for 10+ major product capabilities
- π‘ Led GenAI integration using Azure OpenAI improving user engagement by 40%
- π Reduced execution time by 72% and memory consumption by 63%
- π Increased HCP adoption rates by 38% and model accuracy by 35%
| Project | What | Status |
|---|---|---|
| RAG-based Medical Assistant | Deploying ChromaDB + Mistral 7B medical Q&A to Hugging Face Spaces | π‘ In Progress |
| File Whisperer v2 | Adding multi-doc support and streaming responses | π‘ In Progress |
| Deakin MDS Program | Advanced data science coursework: analytics, modeling, business insights | π’ Active |
| IIIT Bangalore: Agentic AI | Multi-agent systems, LLM orchestration, tool-use patterns | π’ Active |
ποΈ Last updated: June 2026
Your support helps me create more open-source projects and share knowledge with the community.
## β‘ Impact Metrics| Metric | Achievement | Domain |
|---|---|---|
| Performance Optimization | 72% reduction in execution time | Algorithm Engineering |
| Memory Efficiency | 63% decrease in consumption | Enterprise Data Pipelines |
| Business Impact | 38% increase in adoption rates | Predictive Analytics |
| Model Accuracy | 35% improvement in precision | HCP Targeting Models |
| Leadership | Trained 70+ professionals | Python, SQL, Optimization |
| Project Delivery | 25+ successful deployments | Healthcare & Marketing |
| Team Management | Led 5+ data scientists | Cross-functional Collaboration |
| API Architecture | Built Pre/Post-Optimization APIs | System Design & Scalability |
Python PyTorch TensorFlow Scikit-Learn XGBoost Keras
Specializations: Machine Learning β’ Deep Learning β’ Predictive Analytics β’ Statistical Modeling β’ Feature Engineering β’ Time Series Forecasting β’ Computer Vision β’ NLP
Expertise: RAG Systems β’ Prompt Engineering β’ LLM Fine-Tuning β’ Embeddings β’ Semantic Search β’ Inference Optimization β’ LlamaIndex
Docker MLflow GitHub Actions FastAPI Airflow
AWS Azure GCP Databricks Snowflake
SQL PostgreSQL MongoDB Apache Spark
Vector Databases: FAISS β’ Pinecone β’ Weaviate
Tech Stack: Python β’ Optimization Algorithms β’ Azure β’ MLOps β’ SaaS
- Led development of enterprise-scale Marketing Mix Modeling framework for Fortune 500 pharma clients
- Architected 10+ optimization capabilities including Portfolio Optimization, Multi-Level Constraints, and Monthly Gating
- Implemented advanced algorithms (COBYLA, SLSQP, etc.) with non-linear response modeling
- Delivered 25+ MMM projects for Bayer, Merck, Novartis, Janssen with measurable ROI improvements
- Built Pre/Post-Optimization APIs reducing execution time by 72% and memory by 63%
Upload any PDF, DOCX, or TXT and chat with it using AI. RAG-powered document Q&A with FastAPI backend, pgvector semantic search, and Cohere embeddings. Supports BYOK (Bring Your Own Key).
Stack: Python β’ FastAPI β’ LangChain β’ pgvector β’ Cohere β’ React β’ Vercel
Fast, lightweight URL shortener with click analytics, custom aliases, and link expiry. Node.js + Express backend on Hugging Face Spaces, PostgreSQL on Neon, frontend on Vercel.
Stack: Node.js β’ Express β’ PostgreSQL β’ Neon β’ Vercel β’ Hugging Face
Client-side code & text diff tool β paste or upload files, compare with syntax-aware highlighting. Supports Jupyter notebooks, privacy-first (no data sent to server).
Stack: JavaScript β’ HTML5 β’ CSS3 β’ GitHub Pages
Web app converting YAML configurations to Python variable assignments in real-time. Privacy-first, fully client-side processing with js-yaml.
Stack: JavaScript β’ js-yaml β’ HTML5 β’ CSS3 β’ GitHub Pages
End-to-end MLOps pipeline for predicting customer purchase of wellness tourism packages. XGBoost classification with MLflow tracking, Hugging Face data/model versioning, GitHub Actions CI/CD, and Dockerized Streamlit deployment.
Stack: Python β’ XGBoost β’ MLflow β’ Docker β’ GitHub Actions β’ Streamlit β’ Hugging Face
End-to-end MLOps pipeline for engine failure classification using 6 sensor inputs (RPM, oil/fuel/coolant pressure, temperature). MLflow experiment tracking, GitHub Actions CI/CD, and Dockerized Streamlit deployment on Hugging Face Spaces.
Stack: Python β’ Scikit-learn β’ XGBoost β’ MLflow β’ Docker β’ GitHub Actions β’ Streamlit β’ Hugging Face
End-to-end deep learning system for detecting pneumonia from chest X-rays. Trained on the RSNA dataset (26,000+ images) using EfficientNetB3 transfer learning. Supports DICOM and standard image formats with confidence scoring and clinical recommendations.
Model Performance: 74.76% validation accuracy Β· 3-class classification (Normal / Lung Opacity / Not Normal)
Stack: Python β’ TensorFlow β’ EfficientNetB3 β’ CNN β’ Transfer Learning β’ Streamlit β’ Docker β’ Hugging Face Hub
RAG-based medical Q&A over the Merck Manual (19th ed.). ChromaDB semantic search, GTE-large embeddings, Mistral 7B (GGUF) for answer generation. Runs fully locally for privacy with optional GPU acceleration.
Stack: Python β’ LangChain β’ ChromaDB β’ Mistral β’ Sentence-Transformers β’ Jupyter
RAG-powered HR policy Q&A bot for Flykite Airlines employee handbook. Answers employee questions from a PDF knowledge base with page-level citations. Deployed on Hugging Face Spaces with GitHub Actions CI/CD auto-deploy pipeline.
Stack: Python β’ LangChain β’ FAISS β’ Groq (LLaMA 3.3 70B) β’ sentence-transformers β’ Gradio β’ GitHub Actions
Agentic AI chatbot for food-delivery order support. A SQL agent queries a live orders database and an LLM formats the response into natural, empathetic replies. Includes guardrails for blocked queries and automatic escalation to human agents.
Stack: Python β’ LangChain β’ Groq (LLaMA 4) β’ SQLite β’ SQL Agent β’ Gradio β’ GitHub Actions
Marketing Mix Modelling app: attribute sales/revenue to channels with adstock, saturation transforms, and ROI/mROI. Streamlit wizard, 5 model types (Linear, Ridge, Lasso, Bayesian), segment analysis.
Stack: Python β’ Streamlit β’ Scikit-learn β’ Bayesian β’ Optimization
AI-powered MLOps platform that optimizes your resume for Applicant Tracking Systems. ATS scoring, keyword analysis, skill gap insights, and smart job matching.
Stack: Python β’ NLP β’ MLOps β’ Streamlit β’ AI
Interactive roadmap for Data Engineer, Data Scientist, ML Engineer, AI Engineer paths. Progress tracking, clickable topics with resources, study schedules, and interview prep.
Stack: HTML β’ CSS β’ JavaScript β’ GitHub Pages
Free, comprehensive learning platform for mastering Data Science, AI, and ML. 445+ curated problems across 16 topics: Python, ML, Deep Learning, NLP, Computer Vision, and more.
Stack: HTML β’ JavaScript β’ Problem-solving β’ Education
Professional portfolio website: ML/AI projects, Generative AI & MLOps experience, marketing analytics, and product optimization. Apple-inspired design, responsive, FormSubmit contact.
Stack: HTML5 β’ CSS3 β’ JavaScript β’ GitHub Pages
Comprehensive AI & ML project portfolio from University of Texas at Austin PG Program. Real-world data science and machine learning solutions across multiple domains.
Stack: Jupyter β’ Python β’ Scikit-learn β’ Neural Networks β’ MLOps
| Project | Description |
|---|---|
| MDS-Deakin-University | Data science projects from Deakin University MDS program β analytics, modeling, business insights |
| PGP-Applied-AI-Agentic-AI-IIITB | Applied AI & Agentic AI from IIIT Bangalore β LLMs, RAG, multi-agent systems |
| System-Design | System design roadmaps for SDE, ML Engineer, AI Engineer, Data Scientist, Data Engineer |
| Anant-Tripathi | Cyberpunk-inspired portfolio with particle animation |
Manager β Software Engineering (May 2026 β Present)
- Leading software engineering teams delivering scalable AI and analytics products across the Axtria platform
- Driving alignment across product, engineering, QA, and client stakeholder teams for enterprise-grade solutions
- Architected and delivered an in-house Synthetic Data Generation platform accelerating QA testing, UAT cycles, and product demos for major pharmaceutical clients
Project Leader β Data Science / ML (Jun 2024 β Apr 2026)
- Leading 10+ major product capabilities with 95%+ on-time delivery and 98-100% error-free releases
- Architecting scalable optimization systems serving enterprise pharmaceutical clients
- Mentoring team of 5+ data scientists and training 70+ employees
Senior Associate β Data Scientist (Apr 2023 β May 2024)
- Owned MMX optimization enhancements and algorithm implementations (COBYLA, SLSQP, CCSA)
- Led high-impact POCs including Grid Selection, LSTM forecasting, and execution time optimization
- Supported multiple global projects for Novartis brands across Poland and Germany
Associate β Data Scientist (May 2022 β Apr 2023)
- Delivered client-specific enhancements for Janssen and Novartis with custom segmentation
- Designed performance-optimized workflows improving memory utilization significantly
- Researched and validated SLSQP algorithm implementation for Optimization API
Analyst β Data Scientist (Jul 2021 β Apr 2022)
- Built Early Adopter Predictor increasing HCP targeting adoption by 38%
- Delivered 5 Marketing Mix Modeling projects for top US pharma clients
- Established foundation in MMM techniques and analytics workflow delivery
- π Deakin University, Australia | Masters of Data Science (July 2026 β Jun 2027)
- π International Institute of Information Technology, Bangalore | Executive PGP in Applied AI & Agentic AI (Dec 2025 β Aug 2026)
- π The University of Texas at Austin, USA | Post Graduate Program in Artificial Intelligence & Machine Learning (Feb 2025 β Mar 2026)
- π Birla Institute of Technology and Science, Pilani | B.E. & M.Sc. (Integrated) in Electrical and Electronics (Aug 2016 β Jun 2021)
- β
Machine Learning Specialization β Stanford University & Deeplearning.ai (Andrew Ng)
- Comprehensive coursework in supervised/unsupervised learning, neural networks, and ML best practices
- β
Generative AI for Software Developers β IBM
- Practical applications of GenAI in software engineering workflows
- β
Introduction to Generative AI β Google Cloud
- Core concepts and cloud deployment of GenAI solutions
- π Right Brigade Award (Axtria) β Recognized for exemplary display of "RIGHT" values: Responsiveness, Integrity, Get going, Humble, and Team Player
- π Bravo Award (Axtria) β Honored for delivering high-quality work, exemplary performance, and strong client appreciation across multiple high-stakes projects
current_focus = { "research": [ "Agentic AI Systems", "RAG Architectures & Vector Search", "LLM Fine-Tuning & Inference Optimization", "Multi-Agent Coordination" ], "engineering": [ "MLOps Pipelines & Automation", "System Architecture & API Design", "Optimization Algorithms (COBYLA, SLSQP, CCSA)", "Real-time Model Serving" ], "business": [ "Marketing Mix Modeling (MMM)", "Portfolio Optimization", "Product Leadership & Strategy", "Enterprise AI Solutions" ], "learning": [ "Advanced AI/ML Research (UT Austin)", "Applied AI & Agentic Systems (IIIT Bangalore)", "Data Science Mastery (Deakin University)", "Distributed Computing & Cloud Architecture" ], "teaching": [ "Training 70+ professionals", "Technical mentorship", "Knowledge sharing & documentation" ] }
- Azure OpenAI integration and production deployment
- RAG system architecture with vector databases (FAISS, Pinecone, Weaviate)
- Prompt engineering and LLM fine-tuning
- Embeddings and semantic search optimization
- LangChain and LlamaIndex workflows
- Marketing Mix Modeling (MMM) with 25+ delivered projects
- Advanced optimization algorithms: COBYLA, SLSQP, CCSA
- Non-linear response curves (S-curves, diminishing returns)
- Portfolio-level optimization with multi-level constraints
- Budget planning and profit maximization scenarios
- Supervised learning: Random Forest, XGBoost, Logistic Regression
- Time series forecasting and anomaly detection
- Early adopter prediction and HCP targeting
- A/B testing, experiment design, and causal inference
- Model evaluation and hyperparameter optimization
- End-to-end pipeline automation with CI/CD
- Docker containerization and FastAPI deployment
- MLflow for experiment tracking and model versioning
- Cloud deployment: AWS, Azure, GCP, Databricks
- Performance optimization: 72% execution time reduction, 63% memory reduction
Your support helps me create more open-source projects and share knowledge with the community.
I'm always interested in:
- π Collaborating on AI/ML projects
- π‘ Discussing GenAI, LLMs, and optimization strategies
- π Sharing knowledge on MLOps and production ML systems
- π― Exploring opportunities in ML Engineering and AI Research
Reach out:
Email LinkedIn X Portfolio Kaggle
βοΈ From ananttripathi - Building the future of AI, one model at a time